INS NYC 2024 Program

Poster	Poster Session 04 Program Schedule 02/15/2024 12:00 pm - 01:15 pm Room: Majestic Complex (Posters 61-120) Poster Session 04: Neuroimaging \| Neurostimulation/Neuromodulation \| Teleneuropsychology/Technology Final Abstract #78 Inferior Parietal Lobe Activation Predicts Accuracy in a Working Memory Task David Negelspach, The University of Arizona, Tucson, United States Anna Alkozei, The University of Arizona, Tucson, United States William Killgore, The University of Arizona, Tucson, United States Category: Neuroimaging Keyword 1: working memory Keyword 2: neuroimaging: functional Objective: Working memory is a process that involves coordinated engagement among several regions of the brain including frontoparietal networks as well as subcortical regions. While some studies have identified neural correlates that relate to accuracy of working memory, these studies have generally focused on singular regions without side-by-side comparisons. As a result, it is unclear which regions contribute most to working memory performance in healthy individuals. The current study addresses this by comparing activity across areas known to be involved in working memory during functional magnetic resonance imaging (fMRI). Using a whole-brain multiple regression approach, mean-averaged voxel intensity across frontoparietal regions were used to predict response accuracy. Participants and Methods: Twenty-one healthy individuals (8 male, 13 female, mean age = 23.29) with regular sleep schedules (verified by actigraphy) underwent a 3T fMRI while completing an n-back task. Prior to entering the scanner, participants underwent practice trials to ensure they understood how to complete each task. Experimental protocols consisted of 12 trials of n-back sequences. Each trial sequence was separated by 15 seconds of fixation to allow for rest. Trial sequences were randomized across participants. Responses were recorded by a button press following the display of each letter and converted to a total accuracy score. Neuroimaging data were preprocessed and analyzed using SPM 12 in Matlab. Results: The primary contrast, on which group analysis was performed, compared BOLD signal during a two-back task versus periods of fixation. Consistent with previous studies, significant activation was apparent in frontal, parietal, and cingulate cortices. To determine areas of activation most significantly correlated with high accuracy scores, a multiple regression analysis was performed using accuracy scores as the main predictor with age and gender included as covariates. Significant cluster corrected activation was apparent in the inferior parietal lobe (MNI: -36, -40, 44) (P<.05, FDR cluster corrected). Peak voxel intensity in the IPL significantly predicted accuracy scores (β= .84, P<.01) with a moderate effect size (R² = .70). Conclusions: While several regions of the frontoparietal attentional/working memory network are active during an n-back task, not all active regions are significantly associated with response accuracy. The current study suggests that BOLD responses in the inferior parietal lobe (IPL) are the most robust predictor of response accuracy when compared to other active regions. These results suggest that within-subject activation of the IPL region may, with further validation, serve as a robust quantitative representation of working memory performance. Additional investigation is required to confirm if IPL activity can be used to assess the degree of working memory impairment incurred from neurologic or neurobehavioral conditions such as traumatic brain injury, fatigue, or degenerative diseases. Furthermore, these results raise the possibility that interventions such as targeted stimulation of the IPL may potentially improve working memory accuracy.

Poster

02/15/2024
12:00 pm - 01:15 pm
Room: Majestic Complex (Posters 61-120)

Poster Session 04: Neuroimaging | Neurostimulation/Neuromodulation | Teleneuropsychology/Technology

Final Abstract #78

Inferior Parietal Lobe Activation Predicts Accuracy in a Working Memory Task

David Negelspach, The University of Arizona, Tucson, United States
Anna Alkozei, The University of Arizona, Tucson, United States
William Killgore, The University of Arizona, Tucson, United States

Category: Neuroimaging

Keyword 1: working memory
Keyword 2: neuroimaging: functional

Objective:

Working memory is a process that involves coordinated engagement among several regions of the brain including frontoparietal networks as well as subcortical regions. While some studies have identified neural correlates that relate to accuracy of working memory, these studies have generally focused on singular regions without side-by-side comparisons. As a result, it is unclear which regions contribute most to working memory performance in healthy individuals. The current study addresses this by comparing activity across areas known to be involved in working memory during functional magnetic resonance imaging (fMRI). Using a whole-brain multiple regression approach, mean-averaged voxel intensity across frontoparietal regions were used to predict response accuracy.

Participants and Methods:

Twenty-one healthy individuals (8 male, 13 female, mean age = 23.29) with regular sleep schedules (verified by actigraphy) underwent a 3T fMRI while completing an n-back task. Prior to entering the scanner, participants underwent practice trials to ensure they understood how to complete each task. Experimental protocols consisted of 12 trials of n-back sequences. Each trial sequence was separated by 15 seconds of fixation to allow for rest. Trial sequences were randomized across participants. Responses were recorded by a button press following the display of each letter and converted to a total accuracy score. Neuroimaging data were preprocessed and analyzed using SPM 12 in Matlab.

Results:

The primary contrast, on which group analysis was performed, compared BOLD signal during a two-back task versus periods of fixation. Consistent with previous studies, significant activation was apparent in frontal, parietal, and cingulate cortices. To determine areas of activation most significantly correlated with high accuracy scores, a multiple regression analysis was performed using accuracy scores as the main predictor with age and gender included as covariates. Significant cluster corrected activation was apparent in the inferior parietal lobe (MNI: -36, -40, 44) (P<.05, FDR cluster corrected). Peak voxel intensity in the IPL significantly predicted accuracy scores (β= .84, P<.01) with a moderate effect size (R² = .70).

Conclusions:

While several regions of the frontoparietal attentional/working memory network are active during an n-back task, not all active regions are significantly associated with response accuracy. The current study suggests that BOLD responses in the inferior parietal lobe (IPL) are the most robust predictor of response accuracy when compared to other active regions. These results suggest that within-subject activation of the IPL region may, with further validation, serve as a robust quantitative representation of working memory performance. Additional investigation is required to confirm if IPL activity can be used to assess the degree of working memory impairment incurred from neurologic or neurobehavioral conditions such as traumatic brain injury, fatigue, or degenerative diseases. Furthermore, these results raise the possibility that interventions such as targeted stimulation of the IPL may potentially improve working memory accuracy.